Hugging Face's logo

MalyO2
/
working

Object Detection
Transformers
Safetensors
detr
Generated from Trainer
Model card Files
xet
Community
working / README.md
MalyO2's picture
MalyO2
MalyO2/detr_finetune_aug_no_scheduler
e47f917 verified
preview code
|
raw
history blame
5.01 kB
metadata
library_name: transformers
license: apache-2.0
base_model: facebook/detr-resnet-50-dc5
tags:
  - generated_from_trainer
model-index:
  - name: facebook/detr-resnet-50-dc5
    results: []

facebook/detr-resnet-50-dc5

This model is a fine-tuned version of facebook/detr-resnet-50-dc5 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7887
  • Map: 0.55
  • Map 50: 0.6825
  • Map 75: 0.5932
  • Map Small: 0.0
  • Map Medium: 0.5352
  • Map Large: 0.7531
  • Mar 1: 0.1882
  • Mar 10: 0.6735
  • Mar 100: 0.7588
  • Mar Small: 0.0
  • Mar Medium: 0.7158
  • Mar Large: 0.9385
  • Map Object: -1.0
  • Mar 100 Object: -1.0
  • Map Balloon: 0.55
  • Mar 100 Balloon: 0.7588

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 3e-05
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • training_steps: 125
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Map Map 50 Map 75 Map Small Map Medium Map Large Mar 1 Mar 10 Mar 100 Mar Small Mar Medium Mar Large Map Object Mar 100 Object Map Balloon Mar 100 Balloon
2.1236 0.7692 10 1.3396 0.0768 0.1002 0.0897 0.0 0.0966 0.1387 0.0765 0.3735 0.5647 0.0 0.3789 0.9231 -1.0 -1.0 0.0768 0.5647
1.5088 1.5385 20 1.2730 0.1472 0.1875 0.1691 0.0 0.1297 0.2723 0.1059 0.3647 0.6618 0.0 0.5684 0.9 -1.0 -1.0 0.1472 0.6618
1.3182 2.3077 30 1.2273 0.1816 0.2322 0.1918 0.0 0.2368 0.3423 0.1088 0.3941 0.6647 0.0 0.6053 0.8538 -1.0 -1.0 0.1816 0.6647
1.365 3.0769 40 1.0452 0.2476 0.3019 0.2823 0.0 0.3035 0.4146 0.1118 0.4882 0.7559 0.0 0.7158 0.9308 -1.0 -1.0 0.2476 0.7559
1.2013 3.8462 50 0.9825 0.3006 0.3891 0.3233 0.0 0.3747 0.496 0.1324 0.5265 0.7324 0.0 0.6737 0.9308 -1.0 -1.0 0.3006 0.7324
1.3605 4.6154 60 0.9307 0.3655 0.4809 0.4024 0.0 0.3706 0.5922 0.1324 0.5471 0.7294 0.0 0.6684 0.9308 -1.0 -1.0 0.3655 0.7294
1.0117 5.3846 70 0.8867 0.3834 0.5044 0.4222 0.0 0.4086 0.5963 0.1294 0.5882 0.7324 0.0 0.6737 0.9308 -1.0 -1.0 0.3834 0.7324
1.1224 6.1538 80 0.8413 0.478 0.6138 0.5427 0.0 0.472 0.7053 0.1676 0.6265 0.7529 0.0 0.7053 0.9385 -1.0 -1.0 0.478 0.7529
1.0109 6.9231 90 0.8210 0.5281 0.6515 0.5817 0.0 0.5391 0.7497 0.1559 0.6441 0.7735 0.0 0.7316 0.9538 -1.0 -1.0 0.5281 0.7735
1.0771 7.6923 100 0.8153 0.5506 0.6859 0.604 0.0 0.5638 0.7373 0.1794 0.6618 0.7676 0.0 0.7263 0.9462 -1.0 -1.0 0.5506 0.7676
0.9122 8.4615 110 0.7948 0.5551 0.6839 0.6097 0.0 0.5603 0.7503 0.1853 0.6618 0.7824 0.0 0.7526 0.9462 -1.0 -1.0 0.5551 0.7824
0.9918 9.2308 120 0.7887 0.55 0.6825 0.5932 0.0 0.5352 0.7531 0.1882 0.6735 0.7588 0.0 0.7158 0.9385 -1.0 -1.0 0.55 0.7588

Framework versions

  • Transformers 4.46.3
  • Pytorch 2.4.0
  • Datasets 3.1.0
  • Tokenizers 0.20.0